Re: [arts-users] Zeeman effect
Hi Witali,
The line mixing does not matter yet.
OK, not 0/0 but you simply are not having any quantum numbers at all (these
are the the missing numbers)
The nan numbers will matter later. You will see another error if you
fix the first problem. I would suggest you just keep lines with v1=0
anyways, since you won't have any effect from the v1=1 band.
The problem with the missing quantum numbers at the end of the catalog is
there because you are reading a HITRAN .par file. We deprecated quantum
number reading from the .par files because they were not reliable. To
still read raw HITRAN, you can download it again from their
https://hitran.org/lbl/ webpage, selecting the species you want to use
(presumably only O2-66). You then need to modify the parameter options so
that the first entry is the full .par, the second entry is qns', and the
last entry is qns''. I attach a picture of how the webpage selection
looks to me. This way, all species in HITRAN will have all their quantum
numbers read properly. You can load these into ARTS using ReadHITRAN
method again, but setting hitran_type="Online". This will load all the
quantum numbers into ARTS and also set the Zeeman effect from pure
calculations as far as I remember. We maintain a curated version of HITRAN
in the arts-xml-data in case you do not wish to do this. I was hoping to
add line mixing to it down the line but I couldn't find the mixing in
HITRAN online last I checked.
With hope,
//Richard
Ps. picture of HITRAN selection box:
[image: Screenshot from 2020-11-26 12-14-24.png]
Den ons 25 nov. 2020 kl 17:58 skrev :
> Dear Mr. Eriksson
>
>
> Thank you for the answer.
>
> Since i had my script in qpack, I just generated abs_lines in ARTS 2.4 in
> a short way to demonstrate my setup.
>
>
> The code was
>
>
> Arts2{
> abs_speciesSet(species=["O2-Z-66-53.0469e9-53.0869e9","O2-66"])
> VectorCreate(gs)
> ArrayOfQuantumIdentifierCreate(qid)
> ReadXML(gs, "/arts-xml-data-2.4.0/spectroscopy/zeeman/O2-66.g.xml")
> ReadXML(qid, "/arts-xml-data-2.4.0/spectroscopy/zeeman/O2-66.qid.xml")
> ReadHITRAN(abs_lines,"HITRAN2012.par")
> abs_linesSetZeemanCoefficients(abs_lines, qid, gs)
> WriteXML( "ascii", abs_lines, abs_lines.xml")
>
> It seems to have some 'nan' entries. I compared with your example from
> 18.11.2020 and an other difference is that the last 4 numbers are missed.
>
> 52021456295.2042 1.82478855875538e-21 2.82227115907551e-20 61 63 5.622e-10
> 0.00213411290322581 -0.0645374497084855 9172.03671157167 0.71 0 0 0 0.71 0
> 0 9467.90886355786 0.71 0 0 0 0.71 0 0
>
> 52073500265.913 1.02264066638734e-24 5.88280179901669e-20 61 63 5.64e-10
> nan nan 9172.03671157167 0.71 0 0 0 0.71 0 0 9467.90886355786 0.71 0 0 0
> 0.71 0 0
>
> 52542435628.7166 4.0664650946183e-21 2.47600404813118e-20 57 59 5.79e-10
> 0.00241711264367816 -0.0689937941931067 9172.03671157167 0.71 0 0 0 0.71 0
> 0 9467.90886355786 0.71 0 0 0 0.71 0 0
>
> 52594539557.917 2.25794239624139e-24 5.54036634348622e-20 57 59 5.808e-10
> nan nan 9172.03671157167 0.71 0 0 0 0.71 0 0 9467.90886355786 0.71 0 0 0
> 0.71 0 0
>
> 53066952513.2334 8.54258889506482e-21 2.15218198757136e-20 53 55 5.967e-10
> 0.00276410582010582 -0.0741096869563658 9172.03671157167 0.71 0 0 0 0.71 0
> 0 9467.90886355786 0.71 0 0 0 0.71 0 0
>
> 53119086421.6796 4.69902633155893e-24 5.22012465293933e-20 53 55 5.982e-10
> nan nan 9172.03671157167 0.71 0 0 0 0.71 0 0 9467.90886355786 0.71 0 0 0
> 0.71 0 0
>
> 53595786409.1454 1.68833606335819e-20 1.85085761821127e-20 49 51 6.144e-10
> 0.00319595076923077 -0.0800431763685832 9172.03671157167 0.71 0 0 0 0.71 0
> 0 9467.90886355786 0.71 0 0 0 0.71 0 0
>
> Deos line mixing affect this part? I planed to care about that as soon
> as the quantum numbers are fixed.
>
> Best regards
> Witali
> ___
> arts_users.mi mailing list
> arts_users.mi@lists.uni-hamburg.de
> https://mailman.rrz.uni-hamburg.de/mailman/listinfo/arts_users.mi
>
___
arts_users.mi mailing list
arts_users.mi@lists.uni-hamburg.de
https://mailman.rrz.uni-hamburg.de/mailman/listinfo/arts_users.mi
Re: [arts-users] Zeeman effect
Dear Mr. Eriksson
Thank you for the answer.
Since i had my script in qpack, I just generated abs_lines in ARTS 2.4 in a
short way to demonstrate my setup.
The code was
Arts2{
abs_speciesSet(species=["O2-Z-66-53.0469e9-53.0869e9","O2-66"])
VectorCreate(gs)
ArrayOfQuantumIdentifierCreate(qid)
ReadXML(gs, "/arts-xml-data-2.4.0/spectroscopy/zeeman/O2-66.g.xml")
ReadXML(qid, "/arts-xml-data-2.4.0/spectroscopy/zeeman/O2-66.qid.xml")
ReadHITRAN(abs_lines,"HITRAN2012.par")
abs_linesSetZeemanCoefficients(abs_lines, qid, gs)
WriteXML( "ascii", abs_lines, abs_lines.xml")
It seems to have some 'nan' entries. I compared with your example from
18.11.2020 and an other difference is that the last 4 numbers are missed.
52021456295.2042 1.82478855875538e-21 2.82227115907551e-20 61 63 5.622e-10
0.00213411290322581 -0.0645374497084855 9172.03671157167 0.71 0 0 0 0.71 0 0
9467.90886355786 0.71 0 0 0 0.71 0 0
52073500265.913 1.02264066638734e-24 5.88280179901669e-20 61 63 5.64e-10 nan
nan 9172.03671157167 0.71 0 0 0 0.71 0 0 9467.90886355786 0.71 0 0 0 0.71 0 0
52542435628.7166 4.0664650946183e-21 2.47600404813118e-20 57 59 5.79e-10
0.00241711264367816 -0.0689937941931067 9172.03671157167 0.71 0 0 0 0.71 0 0
9467.90886355786 0.71 0 0 0 0.71 0 0
52594539557.917 2.25794239624139e-24 5.54036634348622e-20 57 59 5.808e-10 nan
nan 9172.03671157167 0.71 0 0 0 0.71 0 0 9467.90886355786 0.71 0 0 0 0.71 0 0
53066952513.2334 8.54258889506482e-21 2.15218198757136e-20 53 55 5.967e-10
0.00276410582010582 -0.0741096869563658 9172.03671157167 0.71 0 0 0 0.71 0 0
9467.90886355786 0.71 0 0 0 0.71 0 0
53119086421.6796 4.69902633155893e-24 5.22012465293933e-20 53 55 5.982e-10 nan
nan 9172.03671157167 0.71 0 0 0 0.71 0 0 9467.90886355786 0.71 0 0 0 0.71 0 0
53595786409.1454 1.68833606335819e-20 1.85085761821127e-20 49 51 6.144e-10
0.00319595076923077 -0.0800431763685832 9172.03671157167 0.71 0 0 0 0.71 0 0
9467.90886355786 0.71 0 0 0 0.71 0 0
Deos line mixing affect this part? I planed to care about that as soon as the
quantum numbers are fixed.
Best regards
Witali
___
arts_users.mi mailing list
arts_users.mi@lists.uni-hamburg.de
https://mailman.rrz.uni-hamburg.de/mailman/listinfo/arts_users.mi
Re: [arts-users] Zeeman effect
Hi Witali, The error means that your lines' quantum numbers are defined in a way the ARTS Zeeman module cannot understand (we need F or J to do these computations). If you print the line catalog, you will probably see several "0/0" in there for the lines you have defined as using Zeeman effect. Is this correct? If it is correct, the steps to fix this is to limit where you activate the Zeeman effect, and to define the quantum numbers there correctly. In your "abs_species", however you have defined it, you should limit the Zeeman species to just the few lines you wish to compute it for. This is often done by simply adding "-FLOW-FHIGH" at the end of a species, and then have the species without the "-Z-" defined after the limited species in a different entry. As an example: abs_speciesSet(species=["O2-Z-66-50e9-55e9", "O2-66"]) Now the abs_lines_per_species variable that lbl_checkedCalc takes should have an entry with only the few lines between 50 and 55 GHz. All of these should already have their quantum numbers defined correctly in all standard ways to define them. If you still encounter this problem, you need to manually add the quantum numbers because your original catalog is not well-understood... (Note that you must ensure that you have line mixing defined in your line catalog for O2 16-16 or the above tags will anyways just produce nonsense around the lines.) If it is not correct that you see many 0/0's, please tell me what it looks like and I will try to figure out the problem. With hope, //Richard Den tis 24 nov. 2020 kl 12:43 skrev : > > Hi everyone, > > > I am still working on the Zeeman effect in ARTS 2.4. > > I have set the quantum constants with "abs_linesSetZeemanCoefficients", > but still get the error message: > > > "Run-time error in method: lbl_checkedCalc > > Bad upper state F(s) or J(s)." > > What can that mean? > > Best regards > Witali Krochin > > > > ___ > arts_users.mi mailing list > arts_users.mi@lists.uni-hamburg.de > https://mailman.rrz.uni-hamburg.de/mailman/listinfo/arts_users.mi > ___ arts_users.mi mailing list arts_users.mi@lists.uni-hamburg.de https://mailman.rrz.uni-hamburg.de/mailman/listinfo/arts_users.mi
Re: [arts-users] Zeeman Effect ARTS 2.4 (with qpack)
Hi Witali, The functions in the README require 2 inputs. You need to give them an ArrayOfQuantumIdentifier, qid, and a Vector, gs. The xml-data/spec/zeeman/ contains these files. Load O2-66.g.xml into a Vector. Load O2-66.qid.xml into an ArrayOfQuantumIdentifier. You now call either of the two provided functions with these variables and the data will be loaded into your line-list. To be sure that you have read them in correctly, and that you have done the right thing, save the line-list and look at your lines. They should look something like this: 52021404431.109 1.83322270858126e-21 2.82226897398507e-20 61 63 5.647e-10 0.00213411290322581 -0.0645374497084855 10059.6531675302 0.72 0 0 0 0.72 0 0 10266.7636739206 0.72 0 0 0 0.72 0 0 30 31 31 31 52542394856.9423 4.08755046918299e-21 2.47600265761908e-20 57 59 5.82e-10 0.00241711264367816 -0.0689937941931067 10355.5253195164 0.72 0 0 0 0.72 0 0 10296.3508891192 0.72 0 0 0 0.72 0 0 28 29 29 29 53066912640.8365 8.58174744782781e-21 2.15218139163761e-20 53 55 5.994e-10 0.00276410582010582 -0.0741096869563658 10651.3974715026 0.72 0 0 0 0.72 0 0 10355.5253195164 0.72 0 0 0 0.72 0 0 26 27 27 27 53595757329.277 1.69616777391079e-20 1.8508572209221e-20 49 51 6.173e-10 0.00319595076923077 -0.0800431763685832 10947.2696234888 0.72 0 0 0 0.72 0 0 10414.6997499136 0.72 0 0 0 0.72 0 0 24 25 25 25 The numbers of interest are the two before the first 10k number. Note that you need to have the quantum numbers defined in your line catalog to make this work. In ARTS 2.2 this was optional, I think, but there is an additional and strong J dependency on the Zeeman effect numbers as we showed in a paper from last year: "Updated Zeeman effect splitting coefficients for molecular oxygen in planetary applications". We had to change the method to read these or the code would simply be wrong. Note that the lines above are not going to be directly usable by you. You need to add line mixing. There are other methods in ARTS for that but generally your original line database must be tuned for this to begin with or your forward calculations will make no sense anyways. With hope, //Richard Den mån 16 nov. 2020 kl 18:18 skrev : > > Hi everyone > > I am trying to initialize the Zeeman effect in ARTS 2.4 using the updated > qpack version. > The molecule of interest is oxygen, the frequency range is around > 52-53GHz (the rotational emission lines around 52GHz and 53 GHz). > > I included the method "propmat clearskyAddZeeman" in the propmat clearsky > Agenda. > > However I strugle with setting the quantum numbers and constants. There is > a Zeeman file in the xml Data folder. For me it seems that the required > constants are stored there. The README tell me to use the methods: > > - abs_lines_per_speciesSetZeemanCoefficients or > - abs_linesSetZeemanCoefficients > > I checked the built in documentary, but still not understand how to use > these methods. > For example I don't know in which workspace variable the information > should be feeded in. > > In Arts 2.2 I used readXML to read the file "Zeeman_constants.xml" (was > included in xml data) and feeded the data in the Variable WSMS_AT_START. > > Furhtermore there was an Zeeman demo in the older atmlab version. That was > very helpfull and seems to be missed now. > > Can anyone help me to set this constants? Just a few example lines would > help. > > Best regards > Witali > > > ___ > arts_users.mi mailing list > arts_users.mi@lists.uni-hamburg.de > https://mailman.rrz.uni-hamburg.de/mailman/listinfo/arts_users.mi > ___ arts_users.mi mailing list arts_users.mi@lists.uni-hamburg.de https://mailman.rrz.uni-hamburg.de/mailman/listinfo/arts_users.mi
